In the present study we have selected three different condensation reactions as model reactions, namely the hydroxylalkylation of anisole by paraformaldehyde to bis(methoxyphenyl)methane, the Pechmann condensation of phenols with ethyl acetoacetate (EAA) to coumarins and the Knoevenagel condensation of two aldehydes with three active methylene compounds to form α,β-unsaturated esters and nitriles, using two related Fe-containing metal–organic frameworks (MOFs), namely commercial Fe(BTC) (BTC: 1,3,5-benzenetricarboxylate) and synthetic MIL-100(Fe) as the catalysts. The main aim of this study was to determine the nature of the poisons, the MOF structural stability in connection with the substrate, and the variations in the product selectivity. We have found that undesired intermediates (bisarylmethyl cation in the case of hydroxyalkylation) or byproducts (benzoic acid in the case of Knoevenagel condensation) can poison the MOF by being strongly adsorbed within the MOFs and blocking the pores. In the Pechmann condensation, besides pore blocking, a low structural stability of Fe(BTC) was reflected in the collapse of the crystal structure, while using polyhydroxy aromatic compounds because of their ability to act as ligands for Fe3+, replacing trimesate ligand. MIL-100(Fe) was considerably more robust for this reaction.